Everything this studio costs, and everything it earns. The figures below come straight from the ledger in the repository. No one types them in by hand.
| What | Amount |
|---|---|
| Noah’s hours | 14 h 22 min862 minutes recorded |
| Noah’s time at €100/h | €1,436.67 |
| AI spend | €0.0010 entries, each recorded at zero |
| Service costs | €0.00no entries yet |
| All costs | €1,436.67 |
| Revenue earned | €0.00no entries yet |
| Gifts received | €0.00no entries yet |
| Total income | €0.00no entries yet |
| Running profit and loss | -€1,436.67 |
Day zero — 11 to 31 August 2026
The project began on 1 September 2026. Everything in the three weeks before it counts as day zero: the founding conversation, the specifications, the outside reviews, the website, and the hiring. A man can think about doing a thing for months, and the thinking is never the doing.
Day zero cost real hours and real money, so it is on this page. It is counted separately because the six-month test measures the project, not the run-up to it.
| Noah’s hours | 10 h 51 min651 minutes recorded |
|---|---|
| Noah’s time at €100/h | €1,085.00 |
| Cash spend | €61.14 |
| Day-zero cost | €1,146.14 |
| Everything, both periods | €2,582.81 |
The running profit and loss stands at -€1,436.67, in the red.
The zero against AI spend is real and is not an absence of records. There are 10 AI entries in the ledger and every one of them runs on a subscription that already existed, so the marginal cost of the work was nothing. The sessions are recorded even when the euro figure is zero, because the amount of machine work is part of the story whatever it costs.
Two tests, both published
The cumulative test. All revenue since 11 August 2026, set against all costs since 11 August 2026, my hours included. Day zero is in this number, because that money was really spent. Only this test earns the word profitable.
The run rate. Revenue in the month, set against my time cost in that month. It shows the direction of travel, which is a different question from whether the studio is profitable, and nothing on this page will confuse the two.
Neither number appears without the other.
The cumulative test — the headline
Has the studio earned more since 11 August 2026 than it has cost since 11 August 2026, counting Noah’s hours at €100/h?
Failing. Revenue €0.00 against €2,582.81 of cost. The gap is -€2,582.81, in the red.
Only this test earns the word profitable.
The run rate — trajectory only
Did September 2026 earn more than Noah’s hours cost in September 2026?
Failing. Revenue €0.00 against €1,436.67 of cost. The gap is -€1,436.67, in the red.
This shows the direction of travel. It is not profit and is never reported as profit.
Where gifts go
Every gift is spent on the development itself, whether that means tokens, hardware, or whatever else the work turns out to need. None of it goes anywhere else, and the total sits on this page so you can check.
Gifts are counted here as income, and what they buy is counted as cost. Both appear in the table above, on their own lines, so you can always see how much of the result came from readers rather than from the work.
The word earned is kept for revenue that somebody paid for a thing. A gift is called a gift.
If you want to put something in, the ways to do it are on the support page. A night of machine time costs about 15 euros, and you can sponsor one.
Nothing received so far: no entries yet in the gifts ledger. That is the same claim the table makes about revenue, and it is a different claim from the three figures above that are owed.
How my time is counted
My hours are booked at 100 euros an hour. That is about what it would cost to find someone capable of doing the job instead of me. A studio that counts only cash is telling you it found a way to work for free, and no such way exists.
The synthetic employees
Eleven seats are filled. Each one is a machine given a role, a brief, a perspective, and a job it answers for.
- Chief of Staff. Runs the decision queue and the hiring, and keeps my attention on what only I can settle.
- Game Director. Rules on whether the game remains one thing, and on whether a new decision belongs inside it.
- Lead Engineer. Builds it, and owns whether what shipped works.
- Simulation Director. The world that runs whether or not a player is in it.
- Art Director. How it looks, and whether what came back is good enough to keep.
- QA Director. Whether a passing result counts as evidence at all.
- Player Experience Director. What the game feels like, before anyone has touched it.
- Operations Director. Checks every night that the work happened as the mandate required.
- Staffing Director. Who gets hired into a seat, and on what evidence.
- Marketing Director. Turns verified work into something a reader can use.
- Archive Curator. Keeps the permanent record, which is the studio's memory.
I sit above all eleven as CEO. I set what they work on, I rule on what they built, and I say what publishes.
The reporting lines between the seats are written down, and not one of them has yet carried a real handoff. Eleven seats have files. The studio has run two nights. Those are separate facts, and this page keeps them separate.
Five of the eleven run on claude-opus-5, two on grok-4.6, and one on claude-fable-5. The rest are assigned an engine and have not yet run.
What the target is
Profitable on the cumulative test by 1 March 2027, at five hours of my time a week or less.
The month this project fails that target, the failure will appear on this page in the same size type as everything else.